דוקטורנטים מספרים - Razi Zeidan
Stellar activity characterization using deep learning for extra-solar planet detection
Razi Zeidan (Ph.D. candidate)
Supervisor: Prof. Shay Zucker
It is nighttime. The dark skies are lit by the flickering light of the stars. Your heart skips a beat as you witness this sublime vision and realize the meaning behind it. . Every star is another Sun that can have planets encircling it just like ours. Perhaps alien creatures are looking at our Sun from their home planet the way we are looking at theirs. Perhaps we are not that unique, and Earth is not one of a kind. Perhaps we can somehow detect these habitable worlds.
Fortunately, we live in a special time. We now have the means to detect signals from Sun-like stars induced by Earth-like planets. However, the main obstacle remains unresolved: stellar activity, which can mimic and swamp the planetary signals. Therefore, it is necessary to develop computational tools that can characterize stellar activity and eventually extract planetary signals. This will enable us to discover planets that might support life.
In my research, I use machine learning tools for characterizing stellar activity. It is my hope that eventually it will be possible to detect Earth like exo-planets located in the habitable zone of Sun-like stars more efficiently. These tools are promising and might help us find exo-planets more efficiently on space missions such as TESS and PLATO.
